• Acta Optica Sinica
  • Vol. 41, Issue 7, 0706002 (2021)
Fang Wang* and Jichuan Xing
Author Affiliations
  • Key Laboratory of Photonics Information Technology, Ministry of Industry and Information Technology, School of Optics and Photonics, Beijing Institute of Technology, Beijing 100081, China
  • show less
    DOI: 10.3788/AOS202141.0706002 Cite this Article Set citation alerts
    Fang Wang, Jichuan Xing. Novel Intelligent Long-Distance Optical Fiber Pre-Warning Algorithm[J]. Acta Optica Sinica, 2021, 41(7): 0706002 Copy Citation Text show less
    Structure of long-distance optical fiber pre-warning system
    Fig. 1. Structure of long-distance optical fiber pre-warning system
    Φ-OTDR vibration signal acquisition system
    Fig. 2. Φ-OTDR vibration signal acquisition system
    Vibration event classification process
    Fig. 3. Vibration event classification process
    Structure of the improved neural network
    Fig. 4. Structure of the improved neural network
    Multi-scale decomposition tree of wavelet packet decomposition
    Fig. 5. Multi-scale decomposition tree of wavelet packet decomposition
    ANN network used for classification
    Fig. 6. ANN network used for classification
    DNN network used for classification
    Fig. 7. DNN network used for classification
    Three signal samples in Baoshan area of Shanghai. (a) NI event; (b) MG event; (c) ME event
    Fig. 8. Three signal samples in Baoshan area of Shanghai. (a) NI event; (b) MG event; (c) ME event
    Comparison of signal samples before and after Max Abs Scaler. (a1)(a2) NI event; (b1)(b2) MG event; (c1)(c2) ME event
    Fig. 9. Comparison of signal samples before and after Max Abs Scaler. (a1)(a2) NI event; (b1)(b2) MG event; (c1)(c2) ME event
    Signal and its WPE distributions. (a)NI event; (b)MG event; (c) ME event
    Fig. 10. Signal and its WPE distributions. (a)NI event; (b)MG event; (c) ME event
    Relationship between number of epoch and validation accuracy, training loss during training. (a) Validation accuracy; (b) training loss
    Fig. 11. Relationship between number of epoch and validation accuracy, training loss during training. (a) Validation accuracy; (b) training loss
    Comparison of occurrence probability of MD event in model recognition. (a) Before using the improved neural network; (b) after using the improved neural network
    Fig. 12. Comparison of occurrence probability of MD event in model recognition. (a) Before using the improved neural network; (b) after using the improved neural network
    Type of eventNumber oftraining setsCollection location oftraining sets /kmNumber oftest setsCollection location oftest sets /km
    NI55002.9920002.99
    MG55006.2120006.21
    ME55005.6820005.68
    Table 1. The first experiment data
    Type of eventMethod 1Method 2Method 3
    NI98.295.995.0
    MG97.595.593.1
    ME96.194.292.5
    Average accuracy97.295.293.5
    Table 2. Signal recognition accuracy of the first experimentunit:%
    Type of eventNumber oftraining setsCollection location oftraining sets /kmNumber oftest setsCollection location oftest sets /km
    NI55002.99200014.96
    MG55006.21200016.86
    ME55005.6820006.24
    Table 3. The second experiment data
    Type of eventMethod 1Method 2Method 3
    NI98.095.294.0
    MG96.394.092.1
    ME95.492.790.5
    Average accuracy96.593.892.2
    Table 4. Signal recognition accuracy of the second experimentunit:%
    Number of test setsMethod 1Method 2Method 3
    20001.74 s5.43 s1.88 s
    Table 5. Comparison of the recognition time of the three methods
    Fang Wang, Jichuan Xing. Novel Intelligent Long-Distance Optical Fiber Pre-Warning Algorithm[J]. Acta Optica Sinica, 2021, 41(7): 0706002
    Download Citation